Conditional Technology Moratorium¶
A defined technology activity is paused, or proposed for pause, until a stated safety, evidence, governance, or public-legitimacy condition permits reconsidering its resumption.
Core Idea¶
A conditional technology moratorium is a rule or proposal that suspends a defined technology-related activity while leaving a path to reconsider it when an explicitly stated condition is met. The condition might concern evidence about risk, workable safeguards, a review regime, or public legitimacy. It is not simply “wait”: the activity, the instrument that stops it, the actors reached, and the release condition must be distinguished. A voluntary scientific commitment can bind only its participants; a funding pause controls a funding channel; a legal prohibition, if enacted, would have a different reach. A petition for a moratorium is evidence of a proposed policy form, not proof that anything was actually halted.[1][2][3][4]
The abstraction is a conditional halt-and-release design in science and technology governance. Its conditionality matters because it preserves the possibility of resumption without assuming that time alone makes the activity acceptable. It converts an indefinite dispute into two answerable questions: what exactly is held, and what would have to change to justify lifting the hold? It also exposes a common defect: a proposal can name an ideal such as “consensus” without saying how it will be measured or who may decide that it exists. The proposal still has a conditional form, but its exit pathway may be politically or administratively underdefined.[4]
The original Wikipedia discovery page concerns one narrower case, a proposed ban on developing superintelligence. Its sponsoring statement calls for a prohibition “not lifted before” two conditions, but no government enactment follows from the statement itself. That proposal illustrates the pattern; it is not an exact synonym for every conditional technology moratorium. Earlier examples include selected recombinant-DNA self-restraint and a U.S. federal research-funding pause, whose different instruments demonstrate why “moratorium” cannot be read as a single universal legal ban.[4][1][2]
Structural Signature¶
Sig role-phrases: governed technology activity → suspending instrument and reach → precautionary reason → release condition → evaluation and release pathway → compliance and displacement boundary.
- Governed technology activity. Identify whether the halted object is research, new funding, development, clinical use or deployment, and which subset is covered. A technology name alone is too broad. The 2019 heritable-editing proposal, for example, targeted clinical use, not all editing research.[5]
- Suspending instrument and reach. State what makes the halt operative—voluntary agreement, funder's refusal to support new work, or enacted legal prohibition—and which actors fall within that instrument. Treating a petition as a law or a federal funding pause as a worldwide ban changes the case.[2][4]
- Precautionary reason. A concern about uncertain harms or governance readiness motivates delay despite possible benefits. The concern need not be certified as catastrophic or irreversible in every case; the rule's scope and evidence should be evaluated on their own terms.[1][2]
- Release condition. The suspension is tied to a non-calendar criterion: better risk evaluation, adequate safeguards, adoption of a review policy, or specified scientific and public confidence. Without such a criterion, an indefinite ban or a fixed-date pause is a different design.[1][2][4]
- Evaluation and release pathway. A conference, agency, review process or future political decision may assess the condition and change the hold state. A formal certifying body is not universal; when no decision procedure is specified, the exit is operationally ambiguous rather than conceptually absent.[1][3][4]
- Compliance and displacement boundary. The instrument reaches some actors and not others. It can be followed, evaded or bypassed, but actual defection is not a defining feature. A funding pause leaves activities outside that funding channel untouched unless a separate rule reaches them.[2]
The defining sequence is scope → halt → condition-sensitive reassessment → possible authorized resumption or continued hold. Meeting a necessary condition does not always compel resumption: the 2025 AI statement says a prohibition should not be lifted before two thresholds, without supplying an automatic release rule or certifier. An enacted rule may be more operationally precise. The conditional identity should not be inflated into a guarantee of good governance.[4]
What It Is Not¶
It is not a permanent categorical ban with no contemplated release condition, nor a mere calendar pause whose expiry is independent of evidence or safeguards. It is not ordinary oversight that lets all covered activity proceed while reviews occur; a halt must actually be specified. Conversely, lifting a moratorium into a review framework does not mean every later proposal is approved or that the underlying risk has disappeared.[2][3]
It is not always a government prohibition. The 1974 recombinant-DNA call was voluntary and limited to selected work; the 2014 U.S. action directly paused new federal funding in its declared scope while encouraging others to pause voluntarily. The 2025 superintelligence statement is a call for a prohibition, not an operational ban. Treating all three as the same coercive instrument would erase the very reach condition that determines their effects.[1][2][4]
It is not automatically a Permission-Gated Hold. That live prime requires an otherwise-ready flow blocked until an external authority emits affirmative permission. A self-imposed scientific deferral or a proposal for broad prohibition need not have an otherwise-ready item or a discrete external release token. Nor is a moratorium a Stage Gate Process with multiple stages and escalating commitment, although a review procedure could later be built around one.[2][4]
Scope of Application¶
This pattern appears in technology governance when decision makers regard immediate continuation as harder to justify than a temporary halt while risks, rules or legitimacy are assessed. The governed activity varies: a scientific community can defer a subset of research, a federal funder can suspend new support, a proposed law can restrict development or deployment, and a proposed clinical-use moratorium can leave nonclinical research outside its scope. These cases share conditional delay but not identical authority, enforcement or evidence standards.[1][2][5][4]
The 1975 Asilomar report is especially instructive because it did not recommend that all recombinant-DNA research remain stopped. After a prior limited voluntary deferral, participants recommended that most work could proceed subject to safeguards proportionate to assessed risk, while some work remained inappropriate under then-available protections. Later formal NIH guidelines were another institutional step, not the single event at which all paused research suddenly resumed.[1]
The 2014–2017 U.S. federal funding pause shows an administratively different case: a defined funding channel was held during public deliberation, and NIH announced removal of its pause in December 2017 upon issuance of a departmental review framework. The existence of a condition did not erase case-by-case review after the pause; the new framework governed future funding decisions. The 2025 superintelligence statement is a proposed case whose release conditions are clear in principle but whose enactment, scope test, certifier and enforcement design remain unspecified by the brief statement itself.[2][3][4]
Clarity¶
Four questions separate cases that sound alike. What activity? “AI” or “genome editing” is not enough; development of superintelligence and clinical heritable editing are narrower claims. What instrument? Voluntary self-restraint, withholding public funding, and law have different authority. What condition? A release tied to adopted review rules differs from one tied to scientific consensus and public assent. What status? A signed appeal, a current pause, and a lifted pause are different facts. Answering these before arguing merits prevents both overstatement and false equivalence.[2][3][4][5]
The distinction also clarifies the verb “lift.” A condition can be necessary without being sufficient: “not before consensus” forbids earlier release but need not mandate release the moment consensus is claimed. In contrast, the 2014 government announcement specified adoption of a federal policy as the anticipated end of its funding pause, and the 2017 NIH notice records the actual transition. Conditionality can therefore be an operative rule, a proposal or an underspecified threshold; their statuses should not be merged.[2][3][4]
Manages Complexity¶
An explicit activity boundary and lifting condition reduce a sprawling technology-risk dispute to reviewable decisions. A funder need not decide once and forever whether an entire field is safe; it can identify a subset of new projects whose support is held while a policy is developed. A scientific community can defer a narrow set of work and revise its recommendation when safeguards become credible. This reduction is valuable only when the boundary, evidence and authority remain visible.[1][2]
Too coarse a simplification reverses the benefit. Calling a narrow funding measure a ban on all science conceals permitted activity and exaggerates compliance. Calling an unimplemented petition a moratorium in force confuses advocacy with governance. Conversely, a technology label so broad that no actor can tell whether a particular activity is covered invites inconsistent application. The abstraction manages complexity by exposing these variables, not by pretending they have the same value in every episode.[2][4][5]
Abstract Reasoning¶
To test a proposed case, first write the policy as a conditional rule: for actors in reach \(R\), suspend activity \(A\) while condition \(C\) has not been accepted through pathway \(P\). Determine whether \(A\) is actually stopped, whether \(C\) is more than a date, and whether \(R\) and \(P\) are specified or only aspirational. If the rule concerns only funding, do not infer a ban on privately supported work. If it is a petition, classify it as a proposed moratorium, not an enacted one. If it is a permanent prohibition with no conditional route, the entry's identity fails.[2][4]
Next examine the two errors a conditional hold can make. Releasing too early may expose people to harms the pause was meant to avert; holding too long may defer benefits and encourage work outside the coalition's reach. Neither consequence follows mechanically from the label. The policy analyst must assess whether the specified boundary actually intercepts the relevant risk, whether the evidence threshold can be evaluated, and who has legitimate decision rights to change the status. These are questions generated by the structure, not claims that any historical moratorium succeeded or failed on every dimension.[1][2][4]
Knowledge Transfer¶
The rule transfers across unlike technology settings because it asks the same conditional-governance questions even as the scientific substance changes. The 1974–1975 scientific self-restraint case chose a limited research subset and evaluated safeguards through conference deliberation. The 2014–2017 U.S. case held a funding stream until a formal review regime was adopted. The 2025 AI statement proposes a development prohibition that must not be lifted before scientific and public thresholds, but has no implemented release machinery. One must not copy the evidence standard from one field to another; the transferable part is the halt, its reach and the condition-linked transition.[1][2][3][4]
The portable skeleton might be a general conditional moratorium over any activity, not just technology. That is a future-prime question, not an already established live node. The named entry remains domain-specific because its diagnostic work concerns research/development or use of technology, contested risk evidence and public-governance legitimacy. Its relationship to live Governance and Gatekeeping primes is informative but not strict DAG subsumption under their full definitions.
Examples¶
Scientific self-restraint and the Asilomar transition¶
A 1974 scientific call sought voluntary deferral of certain recombinant-DNA experiments while potential hazards and protective methods were assessed. The 1975 Asilomar conference then considered whether that pause should end. Its original report recommended that most work proceed under safeguards matched to the assessed risk, while retaining caution for a narrower set that could not then be justified under available protections. This was a transition from selected self-restraint to conditional scientific guidance, not a blanket ban lasting unchanged until the later NIH guidelines.[1][6]
The governed technology activity was selected research, not the whole field. The suspending instrument and reach were a voluntary call to researchers rather than worldwide law. The precautionary reason was unresolved potential harm weighed alongside the value of inquiry. The release condition was better understanding and adequate safeguards. The evaluation and release pathway was expert deliberation and risk-differentiated recommendation at Asilomar. The compliance and displacement boundary was the voluntary coalition and the explicit distinction between covered and other research; the source does not prove a particular rate of defection.[1][6]
Mapped back: The case has an identifiable subset, temporary self-restraint and a condition-sensitive move toward resumed work. Its voluntary reach and differentiated release are essential to recognizing the form accurately, without importing later government rules into the earlier pause.
A federal funding pause and framework-based exit¶
In October 2014 the U.S. government announced a pause in new federal funding for a specified subset of life-sciences studies while assessing risks and benefits. It encouraged a broader voluntary pause but did not make that encouragement the same binding instrument as the funding restriction. The announcement linked the pause's end to adoption of a federal policy after deliberation. NIH's December 2017 notice removed its funding pause in connection with the HHS P3CO review framework; future proposals remained subject to review rather than automatically approved.[2][3]
The governed technology activity was federal support for the stated subset of research. The suspending instrument and reach were U.S. funding decisions, not an all-jurisdiction legal prohibition. The precautionary reason was a pending public risk-benefit review. The release condition was a federal policy and review framework. The evaluation and release pathway included deliberation, policy adoption and an NIH notice changing the funding status. The compliance and displacement boundary separated federally funded proposals from other activity outside that funding action's direct control.[2][3]
Mapped back: The same conditional-halt rule appears with a different authority and carrier. The lifted funding pause should not be misreported as either a ban on all work or an unconditional green light for any future study.
A proposed superintelligence-development prohibition¶
The 2025 Statement on Superintelligence calls for a prohibition on development of superintelligence, not lifted before broad scientific consensus that it can be done safely and controllably and strong public buy-in. The statement is an original proposal, not evidence that such a prohibition was enacted. It articulates two necessary release thresholds but does not itself specify a government, operational definition of the covered frontier, certification rule or enforcement regime.[4]
The governed technology activity is the stated superintelligence development target. The suspending instrument and reach are only a proposed future prohibition, with no present legal reach established. The precautionary reason is the proponents' stated concern about severe possible harms. The release condition combines scientific and public thresholds. The evaluation and release pathway remains underspecified in the short statement, a feasibility issue rather than a reason to pretend a certifier exists. The compliance and displacement boundary cannot be reported as actual compliance because the proposal has not been implemented.[4]
Mapped back: This is a clear proposed instance of the conditional form and the original narrow Wikipedia candidate. Its contrast with the two operational histories tests the difference between a rule's conceptual structure and its enacted effect.
Structural Tensions¶
Risk containment versus forgone benefit. Holding risky work can buy time for evidence or safeguards, but delays potentially valuable knowledge or applications; narrowing the halt can preserve benefit yet leave pathways to the feared harm open. Diagnostic: Which covered activity must actually stop to change the claimed risk, and what benefits does that boundary defer?[1][2]
Clear scope versus adaptive coverage. A precise boundary helps actors comply and reviewers judge cases; a quickly changing field can route nearby activities around that boundary. Broader language can catch more activity yet give enforcers and developers less notice of what is permitted. Diagnostic: Can an affected actor classify a proposed activity without relying on ad hoc reinterpretation of the technology label?[2][5]
Demanding release proof versus attainable exit. A high threshold guards against premature resumption, but an unmeasurable consensus or undefined certifier can turn conditionality into de facto indefinite restraint. A weak threshold makes the promise of protection thin. Diagnostic: What observable evidence and decision procedure would establish that the release condition has been met?[4][3]
Coalition legitimacy versus jurisdictional leakage. Voluntary agreement or a funding-only measure can begin quickly and bind willing actors, while leaving activity elsewhere untouched. A wider agreement may cover more risk but requires more coordination and verification. Diagnostic: Which relevant actors fall outside the instrument, and does their activity undermine the pause's own objective?[2][4]
Structural–Framed Character¶
- Evaluative weight: A moratorium embeds a judgment that delay is justified by uncertainty or possible harm; the structure does not itself prove that judgment correct. The benefit-risk balance remains contestable.[1]
- Human-practice dependence: What counts as covered activity, acceptable evidence and legitimate release is negotiated by scientists, funders, publics or lawmakers. No natural process alone confers the policy's binding force.
- Institutional origin: The voluntary Asilomar call, federal funding action and proposed AI prohibition all depend on different institutions or coalitions; no one institution defines the whole pattern.[1][2][4]
- Vocabulary travel: “Moratorium” travels across technologies, but sometimes means a voluntary call, sometimes a funding restriction and sometimes only a proposal. The analyst must carry the instrument and reach with the word.
- Import versus recognition: Recognition requires an actual or proposed halt on specified conduct plus a condition-linked exit. Importing the term onto ordinary oversight, a no-condition ban or a time-limited delay without a release test misses the structure.
Its character: a framed governance rule with a repeatable conditional structure. Its policy meaning is constituted by human authority, risk judgment and legitimacy standards; the abstract halt-and-condition skeleton can be analyzed separately but does not make the named technology-governance entry a prime.
Structural Core vs. Domain Accent¶
The skeletal relation is a condition-bound suspension: specified action is stopped within a declared reach, with change of status dependent on a stated criterion rather than mere elapsed time. A general conditional-moratorium identity beyond technology is an explicit future-prime question, because this package has not established it as an autonomous live prime across unlike nontechnology domains.
The domain-bound mechanism is science and technology governance under uncertainty: research or development boundaries, evidence and safety standards, public legitimacy, actor coalitions and rulemaking authority. The application overlay changes from voluntary scientific self-restraint to federal funding controls or proposed legal prohibition. The named entry does not clear the prime bar because removing the technology-risk and governance setting leaves only the broader conditional-suspension skeleton, not the same field-recognized instrument.[1][2][4]
Instantiates / Related Primes¶
No strict parent edge is proposed in this workspace bundle. Live Governance concerns durable architecture of authority, accountability and decision rights; a single moratorium may be part of such an architecture but can also be a voluntary proposal with no durable binding institution. Live Permission-Gated Hold requires an otherwise-ready flow awaiting external affirmative authorization; selected self-restraint and an unenacted petition do not always meet that exact signature. Live Stage Gate Process requires a sequence of evidence-producing stages and escalating commitments, which is not entailed by one conditional halt. These are analytic neighbors, not lexical parents.
Live Operational Pause is a military capacity-recovery maneuver, unlike technology-risk governance. Live Sunset provision expires a legal authority by date unless renewed; conditional moratorium release depends on a substantive threshold rather than just expiration. A particular case could combine either structure with a moratorium, but that combination cannot be assumed from its name.
Neighborhood in Abstraction Space¶
Conditional Technology Moratorium sits in a sparse region of the domain-specific corpus (85th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.
Family — Sampling, Selection & Accountability Procedures (9 abstractions)
Nearest neighbors
- Timing Risk — 0.82
- Patent map — 0.81
- Requirements analysis — 0.81
- Standard of Care — 0.81
- Rule Against Perpetuities — 0.81
Computed from structural-signature embeddings · 2026-10-08
Not to Be Confused With¶
- Permanent technology ban: lacks an articulated conditional route to reconsider resumption.
- Fixed-duration pause or sunset provision: a date or automatic expiry alone is not a substantive evidence, safety or legitimacy condition.
- Oversight without a halt: may review work while it continues; no moratorium has actually been imposed.
- Funding pause versus legal prohibition: a funder's decision constrains its awards, not all privately or externally supported activity.[2]
- Proposal versus enacted measure: the 2025 superintelligence statement and 2019 heritable-editing call are proposals; they cannot be counted as implemented prohibitions merely because they specify conditions.[4][5]
- All-field research freeze: the 1974 call and 2014 funding action were scoped to subsets, and the 2019 proposal concerned clinical use rather than every research activity.[1][2][5]
- Proof of safety after lifting: transition to safeguards or review changes governance status, not the ontological existence of risk.[1][3]
References¶
[1] Paul Berg et al., “Summary Statement of the Asilomar Conference on Recombinant DNA Molecules”, Proceedings of the National Academy of Sciences 72 (1975), pp. 1981–1984; cited discussion in Introduction and General Conclusions, pp. 1981–1982. Original conference report documenting the limited prior pause and risk-differentiated recommendation; cited only for governance-level facts. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n ↩o ↩p ↩q ↩r ↩s
[2] U.S. White House Office of Science and Technology Policy and Department of Health and Human Services, “Doing Diligence to Assess the Risks and Benefits of Life Sciences Gain-of-Function Research”, 17 October 2014, announcement of funding scope, deliberation and intended policy-based exit. Original government source; no laboratory procedure is used here. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n ↩o ↩p ↩q ↩r ↩s ↩t ↩u ↩v ↩w ↩x ↩y ↩z ↩27
[3] National Institutes of Health, Notice NOT-OD-17-071, “Notice Announcing the Removal of the Funding Pause for Gain-of-Function Research Projects”, 19 December 2017, Purpose and Background. Original notice of transition to framework-based funding review. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k
[4] “Statement on Superintelligence”, hosted by the Future of Life Institute, 2025, statement paragraph and two listed prerequisites. Primary proposal, not an enacted prohibition. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n ↩o ↩p ↩q ↩r ↩s ↩t ↩u ↩v ↩w ↩x
[5] Eric S. Lander et al., “Adopt a moratorium on heritable genome editing”, Nature 567 (2019), proposal and explicit clinical-use scope. Primary proposal, not evidence of a binding universal moratorium. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g
[6] “Chronology of Major Events Associated with Formulation of Policy on Recombinant DNA Molecules”, National Library of Medicine archive, July 1974 entry quoting the contemporary committee call for limited voluntary deferral. registry ↩a ↩b